Elamur Research⌕
☰
home / papers / 2017 / Robust Coreference Resolution and Entity Linking on Dialogues: Character Identification on TV Show Transcripts

Robust Coreference Resolution and Entity Linking on Dialogues: Character Identification on TV Show Transcripts

Summary

The paper addresses character identification in multiparty dialogue. The task links each personal mention (pronouns, nominals, names) in TV show transcripts to the character it denotes. That character may not be a speaker in the scene, and no knowledge base is available. The authors first clean the existing Friends-based corpus of Chen and Choi (2016). They strip stray action notes that had been wrongly included as dialogue. They also split the vague Unknown label into finer categories through manual re-annotation.

For modelling, they build an end-to-end pipeline in two stages:

  • Coreference resolution: an Agglomerative Convolutional Neural Network (ACNN) stacks convolution layers over grouped embedding and discrete features. It produces mention and mention-pair embeddings.
  • Entity linking: a second neural model reuses these representations to form cluster and mention-cluster embeddings, then assigns mentions to characters.

ACNN performs on par with state-of-the-art coreference systems. The cluster-embedding linker clearly beats earlier approaches. The work matters because it shows that coreference modelling tailored to dialogue directly drives entity linking quality. This kind of linking is a building block for dialogue-oriented QA, summarization and assistants.

Key points

  • The corpus covers the first two seasons of Friends, with 18,608 annotated mentions. Former Unknown mentions (about 10% of the annotation) were relabeled as General, Generic or Other over roughly 3 weeks of manual work.
  • ACNN achieves the best B³ scores at both episode and scene level, with comparable results on the other metrics. The comparison is against other coreference systems, including the Harvard system.
  • The episode-level cluster embedding model reaches 86.76% macro F1 and 95.30% accuracy. The mention-embedding-only model already exceeds the heuristic baseline by over 21% F1.
  • On main characters alone, the best model scores 96.38% F1 and 98.42% accuracy. Rare secondary characters (Carol, Mindy, Barry) remain weak, and Mindy makes up only about 1% of the data.
  • The updated corpus and code are public at nlp.mathcs.emory.edu/character-mining/.

Our summary of the paper, in our own words — not its text. The work belongs to its authors; read it at the links above.

Fields: Linguistics & languages

Topics: Named entity recognition · Coreference resolution

Cited by

19 papers of our library cite it (of the 140673 whose bibliographies we read); 4 of them have a page here, 15 not yet.

Cite

The paper itself, not this page. BibTeX from the library's record of the work. DOI 10.18653/v1/k17-1023

@inproceedings{chen2017robust,
  title = {{Robust Coreference Resolution and Entity Linking on Dialogues: Character Identification on TV Show Transcripts}},
  author = {Chen, Henry Y. and Zhou, Ethan and Choi, Jinho D.},
  booktitle = {Proceedings of the 21st Conference on Computational Natural Language Learning (CoNLL 2017)},
  year = {2017},
  doi = {10.18653/v1/k17-1023},
  url = {https://doi.org/10.18653/v1/k17-1023}
}

References

24 references, 9 of them with a card of ours. ↗ leads to the work, «find» searches for one we have no id for yet.

  1. Amit Bagga and Breck Baldwin. 1998. Algorithms for scoring coreference chains. In The first international conference on language resources and evaluation workshop on linguistics coreference. Citeseer, volume 1, pages 563–566. find our card
  2. Shane Bergsma and Dekang Lin. 2006. Bootstrapping path-based pronoun resolution. In Proceedings of the 21st International Conference on Computational Linguistics and 44th Annual Meeting of the Association for Computational Linguistics. Association for Computational Linguistics, Sydney, Australia, pa… find our card
  3. Piotr Bojanowski, Edouard Grave, Armand Joulin, and Tomas Mikolov. 2016. Enriching word vectors with subword information. arXiv preprint arXiv:1607.04606 . find
  4. Yu-Hsin Chen and Jinho D. Choi. 2016. Character identification on multiparty conversation: Identifying mentions of characters in tv shows. In Proceedings of the 17th Annual Meeting of the Special Interest Group on Discourse and Dialogue. Association for Computational Linguistics, Los Angeles, pages… find
  5. Conference of the North American Chapter of the Association for Computational Linguistics on Human Language Technology. NAACL, pages 1020–1030. find
  6. Matthew Henderson, Blaise Thomson, and Steve Young. 2013. Deep neural network approach for the dialog state tracking challenge. In Proceedings of the SIGDIAL 2013 Conference. Association for Computational Linguistics, Metz, France, pages 467–471. http://www.aclweb.org/anthology/W13-4073. ↗
  7. Seokhwan Kim, Luis Fernando D́Haro, Rafael E. Banchs, Jason D. Williams, and Matthew Henderson. 2015. The Fourth Dialog State Tracking Challenge. In Proceedings of the 4th Dialog State Tracking Challenge. DSTC4. find
  8. Xiaoqiang Luo. 2005. On coreference resolution performance metrics. In Proceedings of the conference on Human Language Technology and Empirical Methods in Natural Language Processing. Association for Computational Linguistics, pages 25–32. find our card
  9. Rada Mihalcea and Andras Csomai. 2007a. Wikify!: Linking Documents to Encyclopedic Knowledge. In Proceedings of the Sixteenth ACM Conference on Conference on Information and Knowledge Management. CIKM’07, pages 233–242. find
  10. Kevin Clark and Christopher D. Manning. 2016. Deep reinforcement learning for mention-ranking coreference models. In Proceedings of the 2016 Conference on Empirical Methods in Natural Language Processing. Association for Computational Linguistics, Austin, Texas, pages 2256–2262. https://aclweb.org/a… ↗ our card
  11. Rada Mihalcea and Andras Csomai. 2007b. Wikify!: Linking documents to encyclopedic knowledge. In Proceedings of the Sixteenth ACM Conference on Conference on Information and Knowledge Management. ACM, New York, NY, USA, CIKM ’07, pages 233–242. https://doi.org/10.1145/1321440.1321475. — [doi:10.1145/1321440.1321475](https://doi.org/10.1145/1321440.1321475) ↗
  12. Greg Durrett and Dan Klein. 2013. Easy victories and uphill battles in coreference resolution. In Proceedings of the Conference on Empirical Methods in Natural Language Processing. Association for Computational Linguistics, Seattle, Washington. find our card
  13. Nobal B. Niraula, Vasile Rus, Rajendra Banjade, Dan Stefanescu, William Baggett, and Brent Morgan. 2014. The DARE Corpus: A Resource for Anaphora Resolution in Dialogue Based Intelligent Tutoring Systems. In Proceedings of the Ninth International Conference on Language Resources and Evaluation. LREC… find
  14. Matthew Francis-Landau, Greg Durrett, and Dan Klein. 2016. Capturing semantic similarity for entity linking with convolutional neural networks. In Proceedings of the 2016 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies. Associati… find
  15. Abhishek Gattani, Digvijay S. Lamba, Nikesh Garera, Mitul Tiwari, Xiaoyong Chai, Sanjib Das, Sri Subramaniam, Anand Rajaraman, Venky Harinarayan, and AnHai Doan. 2013. Entity extraction, linking, classification, and tagging for social media: A wikipedia-based approach. Proc. VLDB Endow. 6(11):1126–1… — [doi:10.14778/2536222.2536237](https://doi.org/10.14778/2536222.2536237) find
  16. Stephen Guo, Ming-Wei Chang, and Emre Kiciman. 2013. To Link or Not to Link? A Study on End-to-End Tweet Entity Linking. In Proceedings of the Haoruo Peng, Kai-Wei Chang, and Dan Roth. 2015. A Joint Framework for Coreference Resolution and Mention Head Detection. In Proceedings of the 9th Conference… find
  17. Sameer Pradhan, Alessandro Moschitti, Nianwen Xue, Olga Uryupina, and Yuchen Zhang. 2012. CoNLL-2012 Shared Task: Modeling Multilingual Unrestricted Coreference in OntoNotes. In Proceedings of the Sixteenth Conference on Computational Natural Language Learning: Shared Task. CoNLL’12, pages 1–40. find our card
  18. Lev Ratinov, Dan Roth, Doug Downey, and Mike Anderson. 2011a. Local and Global Algorithms for Disambiguation to Wikipedia. In Proceedings of the 49th Annual Meeting of the Association for Computational Linguistics: Human Language Technologies. ACL’11, pages 1375–1384. find
  19. Lev Ratinov, Dan Roth, Doug Downey, and Mike Anderson. 2011b. Local and global algorithms for disambiguation to wikipedia. In Proceedings of the 49th Annual Meeting of the Association for Computational Linguistics: Human Language Technologies - Volume 1. Association for Computational Linguistics, St… find
  20. Marco Rocha. 1999. Coreference Resolution in Dialogues in English and Portuguese. In Proceedings of the Workshop on Coreference and Its Applications. CorefApp’99, pages 53–60. find
  21. Marc Vilain, John Burger, John Aberdeen, Dennis Connolly, and Lynette Hirschman. 1995. A model-theoretic coreference scoring scheme. In Proceedings of the 6th conference on Message understanding. Association for Computational Linguistics, pages 45–52. find our card
  22. Sam Wiseman, Alexander M. Rush, Stuart Shieber, and Jason Weston. 2015. Learning anaphoricity and antecedent ranking features for coreference resolution. In Proceedings of the 53rd Annual Meeting of the Association for Computational Linguistics and the 7th International Joint Conference on Natural L… find our card
  23. Sam Wiseman, Alexander M. Rush, and Stuart M. Shieber. 2016. Learning global features for coreference resolution. In Proceedings of the 2016 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies. Association for Computational Linguisti… find our card
  24. J. L. Wu and W. Y. Ma. 2017. A deep learning framework for coreference resolution based on convolutional neural network. In 2017 IEEE 11th International Conference on Semantic Computing (ICSC). pages 61–64. https://doi.org/10.1109/ICSC.2017.57. — [doi:10.1109/icsc.2017.57](https://doi.org/10.1109/icsc.2017.57) ↗

↑ back to top

This page in Markdown →